I’d like to thank Regina Fabry for her excellent and detailed response to my paper. She articulates an important account of reading acquisition as a process of enculturation and describes how a Cognitive Integration/Enculturated Cognition (henceforth CI/ENC) account can be combined with a predictive processing account of neural processing. She shows, in impressive detail, how CI/ENC can benefit from Predictive Processing (henceforth PP), primarily as a way of explaining the neural-level details of processes that conspire with bodily interactions with the local environment to complete cognitive tasks. Since Fabry’s response suggests an important way of cashing out some of the details of an enculturated approach, I would like to take this opportunity to look at some of the potential pitfalls in the proposed Enculturated Predictive Processing style (henceforth EPP). Primarily I want to focus on the differences in explanatory emphasis between CI/ENC and PP, especially where CI/ENC proposes the importance of the population-level effects of normative patterned practices (henceforth NPP), such as mathematical practices.
PP is all about predictions, happening in the here-and-now[1]; however CI/ENC occurs at different levels and over much longer time-scales. It turns out that this difference is important, because if the brain is engaged in predictive error minimization (as sub-personal processing) in the here-and-now, then it cannot be driving the innovation of new NPP over many generations. This is because the pressures driving those innovations are found at the social, or populational, level[2], not at the level of neural processing where ‘what?’ is answered in the now.
I also raise several issues concerning the nature of the PP project, particularly whether, as a theory of general brain architecture, all processing can be cashed out in terms of predictive processes. I'm also sceptical about Fabry's claim that PP can provide the “mechanistic underpinnings of the acquisition of cognitive practices” (Fabry this volume, p. 3) on its own, without help from what I call learning-driven plasticity (LDP) and neural redeployment. Finally, I comment on the promising research path down which Fabry is headed.
In the first section I remind the reader of some of the leading ideas of the CI/ENC framework, highlighting, in particular, the different levels of explanation and how this matters to the proposed marriage of ENC-PP. In the second section I raise several problems for the PP approach in general and for the ENC-PP approach in particular. My main concerns are to push away from an ‘isolated brain’ interpretation of PP and to place EPP within a much broader context of explanation.